Gentoo Archives: gentoo-dev

From: Jim Ramsay <lack@g.o>
To: gentoo-dev@l.g.o
Subject: Re: [gentoo-dev] Re: [gentoo-commits] gentoo-x86 commit in www-plugins/adobe-flash: metadata.xml adobe-flash-11.2.202.228.ebuild ChangeLog
Date: Thu, 10 May 2012 17:54:57
Message-Id: 20120510175341.GB16359@freewill.lab.equallogic.com
In Reply to: Re: [gentoo-dev] Re: [gentoo-commits] gentoo-x86 commit in www-plugins/adobe-flash: metadata.xml adobe-flash-11.2.202.228.ebuild ChangeLog by Jim Ramsay
1 On Thu, May 10, 2012 at 01:10:48PM -0400, Jim Ramsay wrote:
2 > On Thu, Apr 26, 2012 at 05:06:45PM -0300, Alexis Ballier wrote:
3 > > wouldnt adding a sse2 useflag and putting it in REQUIRED_USE solve the
4 > > problem ?
5 > >
6 > > afaik portage wont even try to upgrade if people have -sse2
7 >
8 > This is an interesting idea, as it could automatically downgrade users to
9 > 10.3 without requiring any masking. I like that.
10
11 Hmmm... After some testing, I don't think this is how it works.
12
13 It looks to me like if I try to do an upgrade on a system that has
14 USE=-sse2, it doesn't downgrade automatically. Nor does it simply skip
15 the upgrade. At least with portage 2.1.10.57, I get the following fatal
16 output:
17
18 -------
19 # USE=-sse2 emerge -uDN world -p
20 ...
21
22 !!! Problem resolving dependencies for www-plugins/adobe-flash from
23 @selected ... done!
24
25 !!! The ebuild selected to satisfy "www-plugins/adobe-flash" has unmet
26 requirements.
27 - www-plugins/adobe-flash-11.2.202.235::local USE="64bit (multilib)
28 -32bit -kde -sse2 -vdpau"
29
30 The following REQUIRED_USE flag constraints are unsatisfied:
31 sse2
32
33 (dependency required by "@selected" [set])
34 (dependency required by "@world" [argument])
35 -------
36
37 So in light of this, I think I'm going to keep the fatal pkg_pretend
38 with /proc/cpuinfo check in the 11.2 ebuild, since that's the only way I
39 can do the check in a way that meets my 3 requirements:
40 -> No annoyance for users who have sse2 instructions present
41 -> runs at install time in a non-ignorable (fatal) way
42 -> provides useful downgrade instructions to the user
43
44 Too bad pkg_pretend doesn't fire during 'emerge -p' -> That would be
45 even better.
46
47 Any other suggestions on how I can make this work better?
48
49 --
50 Jim Ramsay